Fail to upgrade from 6.13.1 to 7.0.9

[hexregex]

What version are you coming from?

6.13.1

What version are you going to?

7.0.9

What deployment method did you use to deploy?

docker

Did you follow a particular doc? Which one?

yes

Deploy with Docker & Docker Compose

What issues are you running into upgrading?

After upgrading via docker-compose, server load stops on “ServiceBroker message and server remains unavailable” for an infinite time (doesn’t start listen port 3000).

Then we delete current rocketchat database(12Gb size) in mongodb container and restart docker container, rocketchat start normally as expected.

Licence type: CE
Number of users: 159
Server hardware: VM 8CPU/12Gb RAM/60Gb HDD
Version of Rocket.Chat Server: 6.13.1
Operating System: Debian GNU/Linux 12 (bookworm)
Deployment Method: docker
Number of Running Instances: 2(1 instance is for test updates)
DB ReplicaSet OpLog: Enabled
NodeJS Version: 14.21.3 - x64
MongoDB Version: 6.0.13

Before upgrade:

After update to rockerchat:7.0.9 I see:

I spend many hours, reading different issues, try to delete applications from rocket before upgrade, try to delete some indexes in mongosh, try to restore mongodump without indexes, but it’s not help.

After update my nginx have error “104: Connection reset by peer”.
In container I see what node started

/app/bundle # ps a
	PID   USER     TIME  COMMAND
		1 root      0:46 node main.js
	   34 root      0:00 sh
	   48 root      0:00 ps a

but nobody listen port 3000

	docker exec -it fe08fafdf0f0 sh
	/app/bundle # netstat -nlp
	Active Internet connections (only servers)
	Proto Recv-Q Send-Q Local Address           Foreign Address         State       PID/Program name    
	tcp        0      0 127.0.0.11:40089        0.0.0.0:*               LISTEN      -
	tcp        0      0 :::42543                :::*                    LISTEN      1/node
	udp        0      0 127.0.0.11:37436        0.0.0.0:*                           -

I didn’t find any ENV VAR in documentation to set rocketchat to more verbosity log level for catch problem.

[reetp]

Check this first.

By default, GridFS is used in Rocket.Chat for file storage because MongoDB offers this functionality with zero configuration.

However, it is not recommended for production environments due to the high load it places on the database. GridFS can only be suitable if

Look for file migrator scripts to empty your db.

Recommendations for File Upload

[douglascsoliveira]

Apply a 6-month file retention policy.
Migrate to 7.0.0 first

[hexregex]

It’s a lot of work to migrator script(~60 000 files), I need some time to move all file from db to fs to check your idea. We try to clear upload.files & upload.chunks collection, and nothing change rocket 7.0.9 still not start.

[reetp]

@hexregex

It’s a lot of work to migrator script(~60 000 files),

Yup. But it is well documented.

@hexregex

Then we delete current rocketchat database(12Gb size) in mongodb container and restart docker container, rocketchat start normally as expected.

That is telking you something.

You haven’t mentioned number of users or licence type.

Check that against current licencing. A lot changed at 6.5 and more with 7+

Also check github for licence issues - some were more recently fixed.

(All good reasons to read the docs, changelogs & keep up to date)

[hexregex]

@reetp

You haven’t mentioned number of users or licence type.

Check that against current licencing. A lot changed at 6.5 and more with 7+

Also check github for licence issues - some were more recently fixed.

It was In start message(I take this information from admin page on our server):

Licence type: CE
Number of users: 159

But in our personal page on cloud.rocket.chat I see two active workspace with different names:

• rocket-test.domain2.aa self-hosted Enterprise, Yearly active(used by my RC server)
• rocket.domain1.bb self-hosted Community Edition active (this ws don’t used, I can remove it, but my server don’t look at whis ws)

Enterprise licence have status Active(green indicator), but it’s expired and I know it.
On admin page the our RC server I see what we have Community Edition, not Enterprise and the same time in Deploiment page I see what my deploiment use whis expired Cloud Workspace ID.

I can’t to remove expired Cloud Workspace ID and can’t move my server to another new clear CWID. Trying to reenable setup wizard and complete
activation again and again to glue my prod installation with new CWID but it’s not work, it’s only make new workspace i saw it on personal page in cloud.rocket.chat

And the same time this expired WID related with our non prod domain name(maybe it result of some old upgrade experiment on stage server what rewrite right domain name).

After all In rocketchat deployment I still see my old WID

Find this 2 error messages in log then add ENV VAR “LICENSE_DEBUG=true”

{“level”:50,“time”:“2025-04-17T22:25:37.790Z”,“pid”:175067,“hostname”:“rocket.domain1.bb”,“name”:“License”,“msg”:“Url validation failed”,“url”:{“value”:“rocket-test.domain2.aa”,“type”:“url”},“workspaceUrl”:“rocket.domain1.bb”}
Apr 18 01:25:37 rocket.domain1.bb rocketchat[175067]: {“level”:50,“time”:“2025-04-17T22:25:37.791Z”,“pid”:175067,“hostname”:“rocket.domain1.bb”,“name”:“License”,“msg”:“Period validation failed”,“period”:{“validUntil”:“2024-10-15T03:00:00Z”,“invalidBehavior”:“invalidate_license”}}

I write to sales team 2 times from Talk to an expert form but still don’t have any answer message in my mailbox.

Deployment ID
Tuzgz7AAp5PCh7MbP
Cloud Workspace ID
5ea2be02e5358900010f59b2

[reetp]

@hexregex

It was In start message(I

Sorry. Missed that. Too busy.

@hexregex

I write to sales team 2 times from Talk to an expert form but still don’t have any answer message in my mailbox.

Not sure they are particularly bothered about non paying users.

You have reduced yourself to community support.

@hexregex

Period validation failed”,“period”

It will need someone to try & reset the IDs.

I’ll ask but it may take a while - especially as it is Easter.

[hexregex]

@reetp

You have reduced yourself to community support.

It’s not our desigion or some kind of greed. We live in strange time when one morning you become guilty & punished because you were born and live in a certain country.

@reetp

It will need someone to try & reset the IDs.
I’ll ask but it may take a while - especially as it is Easter.

Mayday, Mayday, the clock is ticking, 8 days left until our server turns into a pumpkin.

[reetp]

@hexregex

It’s not our desigion or some kind of greed. We live in strange time when one morning you become guilty & punished because you were born and live in a certain country.

I will refrain from commenting on the politics of your country. I suggest you do the same or you may find you get even less help.

Enterprise licence have status Active(green indicator), but it’s expired and I know it.

So you have knowingly let it expire (because you could have paid to continue it which I know others have) and so you no longer get paid support. That is your choice.

Prior to v7 I believe it required manual intervention to set the IDs and change version. With v7 I believe it now automatically changes.

You have two cloud IDs because you have one for each server. It will require manual intervention to fix.

@hexregex

Mayday, Mayday, the clock is ticking, 8 days left until our server turns into a pumpkin

You can leave your snark where it belongs thanks.

I have asked. It will be what it will be.

I don’t work here and I can’t fix it.

Don’t shoot the messenger.